From: pavenis AT lanet DOT lv Message-ID: To: "Mark E." , djgpp-workers AT delorie DOT com Date: Mon, 23 Aug 1999 17:09:54 +0300 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: Re: gcc flag for binutils 2.10? In-reply-to: <199908221452.OAA20688@out2.ibm.net> X-mailer: Pegasus Mail for Win32 (v3.12a) Reply-To: djgpp-workers AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On 22 Aug 99, at 10:52, Mark E. wrote: > I've been thinking about a way for allow the next version of GCC to take advantage > of the new features in Binutils, at least for the benefit of C++ users. One way > would be to just require Binutils 2.10, but we can't require what isn't released. An > alternative is to add a target specific switch to GCC to allow those with pre- > release versions of Binutils use of the advanced features. > > Here's what I have in mind: > > #define SUBTARGET_SWITCHES \ > { "bnu210", MASK_BNU210, "Enable weak symbols and .gnu.linkonce (requires Binutils > 2.10)." }, \ > { "no-bnu210", -MASK_BNU210, "Disable weak symbols and .gnu.linkonce." }, At least I didn't find MASK _BNU210 in GCC-2.95.1 sources. > This means that -mbnu210 would enable weak symbols and for templates to use > .gnu.linkonce sections. -mno-bnu210 would disable these features if you had them > already enabled (by changing the specs for example). > > The only drawback is you must use the switch with every source file so you don't get > mysterious errors. However, this shouldn't be a problem if you're using RHIDE or a > Makefile. Andris